Downright Dangerous: Decisions Based on Flawed Data
Posted by 
Share this Post
Making decisions based on flawed data can be downright dangerous when it comes to medications
Gary Cokins recently wrote on the topic of analytics being the next differentiator between companies. We covered this topic on this blog a few months back. I especially like Gary's point:
"There is always risk when decisions are made based on intuition, gut feel, flawed and misleading data or politics. In the popular book, Competing on Analytics: The New Science of Winning, author Tom Davenport makes the case that increasingly, the primary source of attaining a competitive advantage will be an organization’s competence in mastering all flavors of analytics. If your management team is analytics-impaired, then your organization is at risk. Analytics is arguably the next wave for organizations to successfully compete and optimize the use of their resources, assets and trading partners."
I completely agree that there is great temptation to base decisions on flawed data, especially if the result is convincing or affirming your original bias. (In the business literature, this is called the anchoring effort, or confirmation bias.)
Let us walk through an example. I recently received a mailer from my health insurance company. It was addressed directly to me as Mr. ABC. I assumed it to be a form letter but a closer look lead me to a different conclusion.
This mailer was sent to me because they noticed that I had missed refill dates on a specific medication I was taking.
The mailer started out with "...it appears that the xxx medication(s) your doctor has prescribed may not be getting filled as often as directed.” They informed me of the potential benefits of this medication, some solutions for common reasons consumers miss filling their prescription and a plug for using generic drugs if cost was the factor bothering me.
(Privacy concerns aside) I am impressed with my health insurance provider's ability to track a specific situation and send me a fairly detailed mailer that has useful recommendations for me. This is possible because of data quality, data analytics and data governance working in conjunction.
For instance, insurance company X might decide to try a pilot project where they proactively help members stay healthy and prevent sickness. To allay privacy concerns they start by inviting interested members to sign up for this pilot program (governance issue). Next they need to track individual ABC's prescription refill frequency amongst other metrics as a measure of health maintenance.
Company X needs to be aware that to suit my needs, I might refill prescriptions at multiple locations (MDM) throughout the city.
Lastly, company X needs to make sure they are getting "high confidence", "high quality" data that can be analyzed (quality + analytics) to generate timely and action inducing mailers.
Mistakes can lead to a range of costly outcomes from privacy lawsuits to members worrying if their claims are not being covered as they thought they were.
To sum up, prevention is indeed better than cure. It saves money for the insurance company and increases quality of life for the member. It is a powerful way to differentiate your brand but you better base the implementation on high quality data gathered through a process that includes appropriate governance structures to ensure a positive outcome.
2 Responses »
Trackbacks
- Tweets that mention Downright Dangerous: Decisions Based on Flawed Data | Mastering Data Management -- Topsy.com
- uberVU - social comments
Leave a Response
Entries(RSS)